Truth and Meaning in#1C8C03

[Pages:29]March 2008

March 2008

Truth and Meaning ? In Perspective

by

Scott Soames

School of Philosophy University of Southern California

To Appear In

Truth and Its Deformities

Edited by Peter French

Midwest Studies in Philosophy Volume XXXII 2008

Truth and Meaning ? In Perspective Scott Soames

My topic is the attempt by Donald Davidson, and those inspired by him, to explain

knowledge of meaning in terms of knowledge of truth conditions. For Davidsonians, these attempts take the form of rationales for treating theories of truth, constructed along Tarskian lines, as empirical theories of meaning. In earlier work1, I argued that Davidson's two main rationales ? one presented in "Truth and Meaning"2 and "Radical Interpretation,"3 and the other in his "Reply to Foster"4 ? were unsuccessful. Here, I extend my critique to cover an ingenious recent attempt by James Higginbotham to establish Davidson's desired result. I will argue that

it, too, fails, and that the trajectory of Davidsonian failures indicates that linguistic understanding, and knowledge of meaning, require more than knowledge of that which a

Davidsonian truth theory provides. I begin with a look at the historical record. The Evolution of an Idea: A Historical Summary

When Davidson enunciated his idea, in the 1960s, that theories of meaning can be taken

to be nothing more than theories of truth, it met with a warm reception. For devotees of Ordinary Language, its attraction lay in its promise of providing a theoretically respectable way

of grounding claims about meaning, and distinguishing them from claims about use, that those who still placed meaning at the center of philosophy had come to recognize the need for.5 For

1 Scott Soames, "Truth, Meaning, and Understanding," Philosophical Studies, 65, 1992; 17-35, and chapter 12 of Philosophical Analysis in the Twentieth Century, Volume 2 (Princeton and Oxford: Princeton University Press), 2003. 2 Donald Davidson, "Truth and Meaning," Synthese, 17, 1967, 304-23; reprinted in Inquiries into Truth and Interpretation, (Oxford: Clarendon Press), 200l. Citations will be to the latter. 3 Donald Davidson, "Radical Interpretation," Dialectica, 27, 1973, 313-28; reprinted in inquiries into Truth and Meaning. Citations will be to the latter. 4 Donald Davidson, "Reply to Foster," in Gareth Evans and John McDowell, eds. Truth and Meaning, (Oxford: Oxford University Press), 1976; reprinted in Inquiries into Truth and Meaning. Citations will be to the latter. 5 See chapters Parts 2-4 of Philosophical Analysis in the Twentieth Century, Vol. 2.

1

those laboring under the Quinean legacy of skepticism about analyticity, synonymy, and meaning, the idea afoot was that extensional notions from the theory of truth and reference were respectable, whereas intensional ones from the theory of meaning were not. This was an audience to which the Davidsonian program was bound to appeal.

It was one thing to claim that meaning has no special role to play in philosophy. As discomforting as this was to Ordinary Language philosophers, it was something that Quineans could live with. Much more troublesome was the idea that meaning had no place in science. It certainly didn't seem that way to soldiers in the Chomskian revolution, who were busy transforming linguistics. The central work of the period, Aspects of a Theory of Syntax,6 enshrined the distinction between deep and surface structure, while championing the thought that a semantic theory of a natural language would interpret the deep structures of its sentences. To many, this brought to mind the Russellian distinction between logical and grammatical form. But what, it was wondered, is logical form, and what would it be to interpret it? Davidson laid the groundwork for answering these questions in a way that made sense to philosophers in the tradition of Russell, Tarski, Carnap, and Quine. For Davidsonians like Gilbert Harman, the logical forms of natural language sentences were their Chomskian deep structures, to interpret them was to give a truth theory for the language, and to see this as a theory of meaning was to see it as explicating what it is to understand the language. 7

Though audacious, these ideas can be seen as the application of a familiar idea from philosophical logic. Since Tarski's seminal work on truth in the 1930s, it has been commonplace to view an interpreted formal language as the result of adding a model, plus a

6 Noam Chomsky, Aspects of a Theory of Syntax, (Cambridge: MIT Press), 1965. 7 Gilbert Harman, "Deep Structure as Logical Form," Synthese, 21, 1970, 275-97.

2

definition of truth-in-a model, to an uninterpreted formal system, thereby arriving at an assignment of truth conditions to every sentence.8 But if truth theories can be used in this way to endow sentences with meaning, then, surely, it seemed, they can also be used to describe the meanings of already meaningful sentences ? provided, in the case of natural language, that we are clever enough to find the requisite logical forms to which to apply them. This was the technical task of the Davidsonian program. The philosophical challenge was to justify the claim that completing this task would yield a theory of meaning.

Coming up with this justification proved to be easier said than done. Consider again the use of a truth theory to endow sentences with meaning. Our announcement that we are using the theory to introduce an interpreted language contains a crucial piece of information not contained in the theory itself ? namely, that certain of its theorems are to be viewed as providing paraphrases of the sentences the truth conditions of which they state. This suggests that if descriptive theories of meaning are to be put in the form of Tarskian truth theories, something beyond what they state must play a crucial role. Also, when we introduce interpreted formal languages, we typically don't have to choose which of the many theorems stating truth conditions of a single sentence provide acceptable paraphrases of it. Since potential paraphrases can often be proved to be extensionally equivalent, each is acceptable for the purposes of philosophical logic, or metamathematics. This is not true when our purpose is to give a descriptive theory of meaning. Thus, if a Tarskian truth theory is to fill the bill, it must be combined with something else that not only provides the information that meaning-

8 Alfred Tarski, "The Concept of Truth in Formalized Languages," and "On the Concept of Logical Consequence," in Logic, Semantics, and Metamathematics, 2nd. Edition, John Corcoran (Indianapolis: Hackett), 1983.

3

giving paraphrases are sought, but also specifies which of the many potential candidates are the genuine articles. This is the heart of the justificatory problem Davidson faced.

Initially, there was widespread optimism about its solution, together with widespread unclarity about what such a solution would require. The optimism was fueled by the attractiveness of the overall picture ? which was seen as applying the proven advances of philosophical logic to the interpretation of natural language, without backsliding on Quine's au currant skepticism about meaning. Davidson thought that systematic knowledge of truth and reference could do all legitimate work for which we need a notion of meaning. His strategy was to embrace Quine's rejection of analyticity, synonymy, and our ordinary notion of meaning, substituting knowledge of truth and reference for it -- whenever there was something genuine to be captured. Since truth and reference are scientifically legitimate, such a theory was deemed respectable. Since it could be used to explain what it is to understand a language, it fit the emerging paradigm in linguistics. In short, was that one can have Quine, and Chomsky too.

There is no need to suppress, of course, the obvious connection between a definition of truth of the kind Tarski has shown how to construct, and the concept of meaning. It is this: the definition works by giving necessary and sufficient conditions for the truth of every sentence, and to give truth conditions is a way of giving the meaning of a sentence. To know the semantic concept of truth for a language is to know what it is for a sentence ? any sentence ? to be true, and this amounts, in one good sense we can give to the phrase, to understanding the language. This at any rate is my excuse for a feature of the present discussion that is apt to shock old hands; my freewheeling use of the word `meaning', for what I call a theory of meaning has after all turned out to make no use of meanings, whether of sentences or of words. Indeed, since a Tarski-type truth definition supplies all we have asked so far of a theory of meaning, it is clear that such a theory falls comfortably within what Quine terms the `theory of reference' as

4

distinguished from what he terms the `theory of meaning'. So much the good for what I call a theory of meaning, and so much, perhaps, against my so calling it.9 There were, however, some conceptual flies in the ointment. The grand Davidsonian-

cum-Quinean theme presupposed that truth and reference can be retained, while meaning is rejected. But it is not clear that our ordinary notions of truth and reference can be separated

from our ordinary notion of meaning. Having rejected meaning as unscientific in Word and Object, and implicitly called ordinary reference into question by implicating it in his

indeterminacy theses, Quine finished it off in "Ontological Relativity," calling for what was, in effect, its elimination.10 Surely, if the ordinary notion of an expression referring to a rabbit is to

be eliminated, then the related notion of a predicate being true of a rabbit must also go. But this brings truth itself into play. How can one hold onto it, once one has abandoned its sister, being

true of? For Quine, the question is moot, since he was willing to trade our ordinary notions of truth and reference for Tarski's disquotational replacements.11

In the beginning, Davidson was too. Initially, he wrongly equated the notion of truth needed in his theories of meaning with Tarski-truth.12 He was also a revisionist about reference.13 He didn't believe that his referential axioms stated facts about the world from

which the truth conditions of sentences follow. On the contrary, these axioms had no independent content, and reflected no independent reality. Rather, they were seen as aspects of

9 Donald Davidson, "Truth and Meaning," p. 24, my emphasis. 10 W.V. Quine, Word and Object (Cambridge: MIT Press), 1960; "Ontological Relativity," in Ontological Relativity and Other Essays (New York: Columbia University Press), 1969. 11 For discussion, see chapter 11 of Philosophical Analysis in the Twentieth Century, plus the final section of my paper, "The Indeterminacy of Translation and the Inscrutability of Reference," Canadian Journal of Philosophy, 29, 1999, 321-70. 12 See my "Tarski's Theory of Truth," Journal of Philosophy 81, 1984, 411-29; and pp. 102-07, and 238-44 of my Understanding Truth (New York: Oxford University Press), 1999. 13 See, Donald Davidson, ""Reality Without Reference," Dialectica, 31, 1977, 247-53.

5

the total theory that derive their content entirely from their role in connecting theorems about the truth-conditions of sentences with one another. On this picture, one derives a statement of the truth conditions of S from statements about the reference of S's parts. The contents of these referential statements are abstracted from their role in deriving statements about the truth conditions of other sentences containing those parts. But since those other sentences contain additional words not in S, further referential axioms are required to derive theorems stating their truth conditions, thereby linking their interpretations to that of S. And so it goes, until the contents of every sentence and word are intertwined with, and dependent upon, the contents of every other sentence and word. In the end, Davidson thought, our understanding any word or sentence is conceptually dependent on our understanding of every other word and sentence ? a radical version of meaning holism, akin to Quine's own.

We decided a while back not to assume that parts of sentences have meanings except in the ontologically neutral sense of making a systematic contribution to the meaning of the sentences in which they occur. Since postulating meanings has netted nothing, let us return to that insight. One direction in which it points is a certain holistic view of meaning. If sentences depend for their meaning on their structure, and we understand the meaning of each item in the structure only as an abstraction from the totality of sentences in which it features, then we can give the meaning of any sentence (or word) only by giving the meaning of every sentence (and word) in the language. Frege said that only in the context of a sentence does a word have meaning; in the same vein he might have added that only in the context of a language does a sentence (and therefore a word) have meaning.14 That was the grand philosophical canvass on which Davidson painted. Some of its main elements, like the idea that Tarskian truth predicates can be used for Davidson's purposes

14 Davidson, "Truth and Meaning," p.22, my emphasis.

6

were just errors to be recanted.15 Other parts ? like Quine's critique of analyticity and synonymy, and his indeterminacy theses ? are vulnerable to powerful objections.16 Although the Davidsonian program continues to this day, much of the original philosophical background for it has fallen away. I will not, therefore, presuppose it in what follows. Instead, I will freely substitute the ordinary notion of truth for Tarski's, I will not assume that intensional semantic notions are illegitimate, and I will not rely on any kind of semantic holism. My question is, How, if at all, can one justify Davidson's claim that theories of truth qualify as theories of meaning? I will approach this question with as little philosophical baggage as possible. The Problem of Justification

Davidson originally held that a truth theory for L qualifies as a theory of meaning, if knowledge of what it states is sufficient for understanding L. The problem was in showing that his theories satisfied the condition. How can knowledge of a truth theory be sufficient for understanding meaning, when its theorems give truth conditions of sentences only in the weak sense of pairing them with materially equivalent claims? If all I know about S is expressed by the theorem `S' is true iff P, I can readily draw the conclusions expressed by S doesn't mean that ~P and S doesn't mean that Q, where the claim made by Q is obviously incompatible with that made by P. But how does one move from these modest negative results to interesting positive conclusions about what S does mean? Initially, Davidson thought that compositionality gave the answer. In compositional theories, theorems stating the truth conditions of sentences are derived from axioms interpreting their parts. Thus, he reasoned, "accidentally true"

15 See pp.. 422-24 of my "Tarski's Theory of Truth,", and pp. 102-07 of Understanding Truth. Also, D. Davidson, "The Structure and Content of Truth," (The Dewey Lectures 1989), Journal of Philosophy, 87, 1990, 279-328. 16 See Philosophical Analysis in the Twentieth Century, chapters 16 and 17 of Vol. 1, chapters 11 and 12 of Vol. 2.

7

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download